Recombinant and natural cellular assays for human G-protein-coupled receptors are used to optimize initial lead molecules obtained from screening. Although the activity of these molecules can be assessed on human genotype receptors, there is increasing evidence that cells impose a phenotypic selectivity to molecules in various cellular backgrounds. This opens the possibility of dissimulations between activity seen in lead optimization assays and the intended therapeutic value in humans. This review discusses the mechanisms by which cells can impose phenotypic selectivity on molecules and approaches to reduce this practical problem for drug discovery.